[
https://issues.apache.org/jira/browse/HDDS-10409?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17825015#comment-17825015
]
Devesh Kumar Singh commented on HDDS-10409:
-------------------------------------------
Since Recon has no way to know when to stop tracking any DEAD or decommissioned
node , Recon will continue to list DEAD or decommissioned nodes forever.
Because sometimes DEAD or decommissioned nodes may come back and rejoin
cluster. In some cases, it is being observed that some DNs being formatted and
wipes out complete metadata of cluster and then recommissioned back to rejoin
cluster with different UUID, so there is absolutely no way for ozone/recon to
know if it is same node and to make things simpler, it is being suggested that
explicit removal of DEAD or decommissioned datanode should be supported
whenever admin user acknowledges to stop tracking such datanodes.
> Decommissioning of datanodes - Duplicate ozone nodes display as dead in
> Ozone Recon
> ------------------------------------------------------------------------------------
>
> Key: HDDS-10409
> URL: https://issues.apache.org/jira/browse/HDDS-10409
> Project: Apache Ozone
> Issue Type: Bug
> Components: Ozone Recon
> Affects Versions: 1.5.0
> Reporter: Devesh Kumar Singh
> Assignee: Devesh Kumar Singh
> Priority: Major
>
> Decommissioning of datanodes - Duplicate ozone nodes display as dead in
> Ozone Recon:
> Use Case: After decommissioning and reformatting a datanode, Recon should be
> able to learn that a datanode was decommissioned from SCM.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
